In partnership with
Flow time: 5 min I your weekly pulse on AI news, tool and case studies reshaping the water sector
The Simplest Way to Create and Launch AI Agents and Apps
You know that AI can help you automate your work, but you just don't know how to get started.
With Lindy, you can build AI agents and apps in minutes simply by describing what you want in plain English.
→ "Create a booking platform for my business."
→ "Automate my sales outreach."
→ "Create a weekly summary about each employee's performance and send it as an email."
From inbound lead qualification to AI-powered customer support and full-blown apps, Lindy has hundreds of agents that are ready to work for you 24/7/365.
Stop doing repetitive tasks manually. Let Lindy automate workflows, save time, and grow your business
🔍 What’s in today’s flow
🌊 AI ocean robots from Tethys mimic sea creatures to explore underwater environments safely and autonomously
☀️ Google’s Project Suncatcher plans to power AI data centers in space, cutting carbon and water use on Earth
🌪️ DeepMind’s Weather Lab outperformed traditional models, forecasting hurricanes up to 15 days ahead
🧠 Kosmos AI acts as a digital scientist, reading thousands of papers and linking every conclusion to source data
⚠️ Shadow AI poses new cybersecurity risks as companies deploy unmonitored AI-generated code across systems
🔬AI research spotlight: Ocean robots that navigate like sea creatures
Source: Tethys
The details
A new generation of AI-powered underwater robots from Tethys Robotics is changing how we explore and monitor the underwater world. Built as compact drones (~35 kg) that can dive to about 300 metres, these robots use an advanced sensor-fusion “brain” that mimics sea creatures: combining acoustic sensing (like dolphin-style echolocation), magnetic and inertial sensing (similar to how fish and turtles orient to Earth’s field and motion), and cameras to build a real-time 3D understanding of their surroundings even in dark, murky water where GPS and normal vision fail/
Key points
The robots support both remote control and fully autonomous modes, planning routes, avoiding obstacles, and collecting data with human override when needed.
Each unit is small and quickly deployable (two people, ~10 minutes), reducing the need for large vessels and complex logistics.
In pilots, they’ve been used to inspect offshore wind assets and subsea infrastructure, and to detect and classify underwater munitions while keeping people out of hazardous waters.
Why it matters
For water utilities, regulators, and coastal managers, this kind of AI-enabled “robot diver” can transform how we inspect, protect, and understand aquatic environments. Over time, fleets of such robots could give the water sector a rich, high-resolution picture of what’s happening below the surface, supporting better asset management, environmental compliance, and long-term ecosystem health.
🤖Latest in AI: Scaling AI with Solar Power from Space
Source: googlestudio.com
Google has unveiled Project Suncatcher, a moonshot idea to run AI data centers in space using constellations of solar-powered satellites equipped with Google TPUs and ultra-high-speed optical links. In constant sunlight, these satellites could generate up to eight times more solar power than panels on Earth, potentially offering nearly continuous, low-carbon energy for AI workloads. Early tests show Google’s AI special computer chip can withstand the radiation of low Earth orbit, and a first trial mission with two prototype satellites is planned for 2027 to validate the concept.
Why it matters
By moving AI infrastructure off Earth, Project Suncatcher could reduce the massive water demand of traditional data centers used for cooling. This means cleaner, faster, and more efficient AI that supports climate forecasting, smart water management, and real-time monitoring , all powered by renewable energy from space.
🔧 Case study: Google DeepMind’s AI for Hurricane Forecasting
Source: KXANews.com KXAN News
Google DeepMind’s Weather Lab used AI trained on decades of cyclone data to predict storm formation, path, and intensity up to 15 days ahead - much longer than traditional models. During the 2025 hurricane season, it worked alongside the U.S. National Hurricane Center and outperformed top weather models, giving faster and more accurate forecasts.
What happened
For the 2025 Atlantic hurricane season, Weather Lab’s AI forecasts were provided in real time to the U.S. National Hurricane Center (NHC) alongside traditional physics-based models. Early evaluations by independent researchers and media reviews show that the AI system consistently outperformed leading numerical models, including the U.S. Global Forecast System, on both track and intensity forecasts.
Why it matters
Better storm forecasts help water utilities and coastal managers prepare earlier, from securing desalination plants and dams to managing flood risks and stormwater systems. With more lead time and reliable data, water operators can protect infrastructure, reduce pollution from overflows, and keep communities safer during extreme weather.
🔧Trending tool: Kosmos – AI Scientist for Autonomous Discovery
Source: Eddison.com
Kosmos is a next-generation AI research assistant that reads, analySes, and connects insights from thousands of scientific papers to uncover new discoveries.
Key features
Reads and organises information from ~1,500 research papers in a single run, plus tens of thousands of lines of analysis code.
Uses a “structured world model” so it can stay focused on one research goal over very long reasoning chains (tens of millions of tokens). Has already reproduced and generated new scientific findings across neuroscience, materials science,
Designed for full traceability: every conclusion in a report links back to specific code and paper passages, making results auditable.
⚖️ AI Tool Scorecard
Ease of use: ⭐⭐⭐☆☆More like a deep-research lab tool than a simple chatbot; it takes time to learn how to prompt it well, and the team notes there are still some UI “rough edges.”
Cost: ⭐⭐☆☆☆Priced betwe
positioned as something you use on high-value questions. There is a generous free tier for academics, but overall it’s expensive for casual
Security & privacy: ⭐⭐⭐ ☆☆Public material focuses more on scientific validity and transparency than on detailed security/privacy guarantees. Organisations with strict data policies will need to review terms and data-handling docs directly.
Integration: ⭐⭐⭐ ☆☆ Vailable via the Edison platform today, but current information suggests a focused research environment rather than broad plug-and-play integrations with many external tools (at least for now).
Overall: 11/20 - It stands out for its transparency, linking every conclusion back to original data and code, making its results easy to verify. While it offers immense value for research teams tackling complex scientific problems, its high cost and learning curve make it less suitable for everyday or small-scale users.
🕵️AI’s shadows: ‘Invisible’ code and rising security risks
A new report from security firm Cycode warns that organisations are already in a “Shadow AI” crisis. Almost every company surveyed has AI-generated code running in its systems, yet 81% of security teams don’t know exactly where or how AI is being used, and 65% say AI has increased their security risk. In many places, developers are using AI coding tools without clear rules, tracking, or governance, leaving unmanaged code and models scattered across the software supply chain.
Takeaway
Unchecked AI use in coding can endanger critical infrastructure, including water utilities, where unseen flaws could disrupt plant operations or expose sensitive data. Organisations must track, test, and govern AI-generated code before it becomes a silent security threat.
Thanks for reading! I hope you’ve enjoyed this week’s edition and look forward to seeing you next week!
Dr. Andrea G.T